RESEARCH ARTICLE

Learning context modulates aversive taste strength in honey beesMaria Gabriela de Brito Sanchez1,2,*, Marion Serre1,2, Aurore Avargues-Weber1,2, Adrian G. Dyer3 andMartin Giurfa1,2

ABSTRACTThe capacity of honey bees (Apis mellifera) to detect bitter substancesis controversial because they ingest without reluctance different kindsof bitter solutions in the laboratory, whereas free-flying bees avoidthem in visual discrimination tasks. Here, we asked whether thegustatory perception of bees changes with the behavioral context sothat tastes that are less effective as negative reinforcements in a givencontext become more effective in a different context. We trained beesto discriminate an odorant paired with 1 mol l−1 sucrose solution fromanother odorant paired with either distilled water, 3 mol l−1 NaCl or60 mmol l−1 quinine. Training was either Pavlovian [olfactoryconditioning of the proboscis extension reflex (PER) in harnessedbees], or mainly operant (olfactory conditioning of free-walking bees ina Y-maze). PER-trained and maze-trained bees were subsequentlytested both in their original context and in the alternative context.Whereas PER-trained bees transferred their choice to the Y-mazesituation, Y-maze-trained bees did not respond with a PER to odorswhen subsequently harnessed. In both conditioning protocols, NaCland distilled water were the strongest and the weakest aversivereinforcement, respectively. A significant variation was found forquinine, which had an intermediate aversive effect in PERconditioningbut amore powerful effect in the Y-maze, similar to that of NaCl. Theseresults thus show that the aversive strength of quinine varies with thelearning context, and reveal the plasticity of the bee’s gustatorysystem. We discuss the experimental constraints of both learningcontexts and focus on stress as a key modulator of taste in the honeybee. Further explorations of bee taste are proposed to understand thephysiology of taste modulation in bees.

KEY WORDS: Gustation, Learning, Pavlovian conditioning, Operantconditioning, Negative reinforcement, Apis mellifera

INTRODUCTIONTaste is the sense that allows animals to distinguish betweenchemical compounds and the sensations they produce based oncontact chemoreceptors. It allows the discrimination of edible fromnon-edible items because the latter have usually a bitter taste(Yamamoto et al., 1994; Scott, 2004; Reilly and Schachtman, 2008;Yarmolinsky et al., 2009). Animals rapidly learn to avoid bittersubstances and, as a consequence, numerous learning protocols useaversive tastes as reinforcers to promote robust learning andmemory (Darling and Slotnick, 1994; Laska and Metzker, 1998;Ito et al., 1999; Gerber and Hendel, 2006; Kemenes et al., 2011;Salloum et al., 2011).

In the honey bee Apis mellifera, an insect that has amodel status forthe study of learning andmemory (Menzel, 1999;Menzel and Giurfa,2001; Giurfa, 2007; Avargues-Weber et al., 2011; Giurfa and Sandoz,2012), gustatory perception has been less well studied (de BritoSanchez et al., 2007; de Brito Sanchez, 2011). Despite the biologicalrelevance of taste for these insects given their contact with andcollection of different types of nectar and pollens, resins, water andother natural products (deBritoSanchez et al., 2007; deBrito Sanchez,2011), less is known about their taste discrimination abilities. Forinstance, the capacity of bees to detect bitter substances iscontroversial (de Brito Sanchez et al., 2005; de Brito Sanchez et al.,2014). In fact, the use of the term ‘bitter’ can be questioned in the caseof the honey bee because it is not knownwhether this insect perceivesquinine and other chemical components as bitter.Here andhenceforth,we use the term ‘bitter’ as in the insect gustatory literature, withoutmaking claims on perceptual sensations (e.g. Tanimura and Kikuchi,1972; Glendinning et al., 2001, 2002; Weiss et al., 2011).

While harnessed bees in the laboratory ingest without reluctancedifferent kinds of noxious substances, including bitter ones (e.g.quinine, salicin, amygdaline and L-canavanine) and even die as aconsequence of the malaise induced by this ingestion (Ayestarán et al.,2010), free-flying bees avoid bitter substances used as a penalty invisual discrimination tasks (Chittka et al., 2003; Avargues-Weber et al.,2010;Rodríguez-Gironés et al., 2013). The first scenario corresponds tothe typical preparation used in the laboratory to study Pavlovianlearning in bees: the olfactory conditioning of the proboscis extensionreflex (PER), an appetitive reflex triggered by sucrose solution. Insectsare harnessed individually in tubes, left for several hours in a restingsituation and then exposed to an odorant (the conditioned stimulus orCS) paired with sucrose solution (the unconditioned stimulus or US)(Takeda, 1961; Bitterman et al., 1983; Giurfa and Sandoz, 2012).Harnessed bees learn to associate the odor with the appetitive sucrosesolution and exhibit afterwards PER to the conditioned odor. It thusseems that under these harnessed conditions, bitter and other aversivesubstances (e.g. concentrated NaCl solution), which are actually toxicfor the bees (Ayestarán et al., 2010), are either not detected or detectedand ingested because of lower acceptance thresholds.

The second scenario corresponds to another typical preparationused to study visual learning in bees: free-flying bees are trained tochoose a visual target associated with sucrose reward and to avoid adistracter associated with the absence of reward or with an aversivesubstance. The associations built in these contexts can be eitheroperant, classical or both, i.e. they may link the response of theanimal (e.g. landing) and the reward/punishment (US), the visualstimuli (CS) and the US, or both. The experimental framework isnevertheless mainly operant because the bee’s behavior isdeterminant for either obtaining the sucrose reinforcement or not(Avargues-Weber et al., 2011). In this context, quinine solutionacts as a negative reinforcement, improving the performances ofvisual discrimination in protocols where a color is associated withsucrose, and a distracter with water (neutral) or quinine (bittersubstance) (Chittka et al., 2003; Avargues-Weber et al., 2010;Received 25 November 2014; Accepted 19 January 2015

1University of Toulouse, Research Center on Animal Cognition, Toulouse 31062,Cedex 9, France. 2CNRS, Research Center on Animal Cognition, Toulouse 31062,Cedex 9, France. 3School of Media and Communication, RMIT University,Melbourne, Victoria 3000, Australia.

*Author for correspondence (maria.de-brito-sanchez@univ-tlse3.fr)

949

© 2015. Published by The Company of Biologists Ltd | The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

Rodríguez-Gironés et al., 2013). Compared with water, quinineimproves the bees’ visual discrimination, leading to the suggestionthat its penalizing nature increases attentional processes(Avargues-Weber et al., 2010; Avargues-Weber and Giurfa,2014). It thus seems that under free-flying conditions, bittersubstances are detected and rejected, contrary to what occurs inharnessed conditions.Here, we raised the question of whether gustatory perception of

honey bees changes with the behavioral context so that aversivetastes that are less effective as negative reinforcements in a givencontext become more effective in a different context. We trainedbees to discriminate one odor (CS+) paired with sucrose fromanother odor (CS−) paired with either distilled water (a neutralsubstance), NaCl or quinine solution (both toxic substances).Conditioning was either Pavlovian (olfactory conditioning of PER),or mainly operant (training of free-walking bees to discriminate thesame two odors within a smallY-maze). PER-trained bees were thentested both in their original context and in the Y-maze whereasmaze-trained bees were tested both in their original maze contextand in the Pavlovian PER context. In this way, we aimed todetermine whether the evaluation of taste changes depending on thebehavioral (Pavlovian, operant) context, i.e. whether aversive tastesare perceived differently by harnessed and free-walking bees.

RESULTSFour experiments were performed (Fig. 1C). In experiment 1,harnessed bees were subjected to a differential olfactory PERconditioning (Fig. 1A) in which they had to learn to discriminatebetween an odor rewarded with sucrose and a non-rewarded odorpaired with water, salt or quinine; retention was afterwardsmeasured in the operant context of a small Y-maze in which theycould freely walk and choose/avoid the two odors previously trained(Fig. 1B). In experiment 2, training was identical to that ofexperiment 1 but retention was measured in the same Pavlovian

context and at the same time as the operant testing of experiment 1(Fig. 1B). This experiment was conceived as a control of experiment1, to verify that performance variation between contexts ofexperiment 1 was due to context change and not to an unspecificmemory decay. It was thus important to determine whether 2 h afterPER conditioning, and after being subjected to the same post-conditioning handling, bees conserved the memories acquired in theolfactory PER conditioning when tested in the same context. Inexperiment 3, training occurred in the Y-maze and retention wasafterwards measured in the Pavlovian context as each bee wasindividually harnessed and tested for PER to the odors previouslyconditioned in the maze (Fig. 1C). Finally, in experiment 4,conceived as a control for experiment 3, both conditioning andtesting occurred in the operant context of the Y-maze. As forexperiment 2, the goal of experiment 4 was to determine whetherperformance variation in experiment 3 was due to a nonspecificmemory decay or to the change of contexts. Importantly, inswitching bees from one experimental context to the other, coolingwas never used to immobilize them as it has amnesic effects, whichcould potentially impair the memories acquired in the first context(Menzel et al., 1974; Erber et al., 1980). Instead, bees were carefullyhandled with individual forceps whose tips were covered with softfoam rubber.

Experiment 1: differential conditioning in the Pavloviancontext of PER conditioning and retention in the operantcontext of the Y-mazeAcquisition in Pavlovian PER conditioningBees were conditioned during 12 trials to discriminate two similarodors, 1-nonanol and 1-octanol, one of which (CS+) was pairedwith an appetitive reward of sucrose 1 mol l−1 (6 CS+ trials) and theother (CS−) with distilled water (water group), quinine 60 mmol l−1

(quinine group) or NaCl 3 mol l−1 (NaCl group) (6 CS− trials).Within each of these groups (water, quinine and NaCl), there were

Experiment 1

Experiment 2

Experiment 3

Experiment 4

Train TestA B C

Fig. 1. Pavlovian olfactory conditioning of harnessed bees andoperant conditioning of free-walking bees in the Y-maze set-up. (A) Top panel shows a beeimmobilized in a metal tube facing an olfactory stimulation device controlled by a computer. The toothpick soaked in sucrose solution allows delivery of sucrosereward to the antennae and mouthparts. Bottom panel illustrates the proboscis extension reflex (PER). (B) Y-maze experimental set-up. Top view of the acrylicY-maze used for conditioning bees in an olfactory discrimination task. Each beewas transported to the entrance zone of themaze, where it was released. The beemoved towards the decision area, delimited by the dashed lines on the figure, where it had to choose between the two odors. The airflow ensured odor diffusion.Odor detection at the decision area and/or arm entrancewas followed by the reinforcement assigned to each odor (sugar solution on one side andNaCl, quinine orwater on the other). Owing to the spatial arrangement of odor and reinforcement, bees experienced first the odor and then the reinforcement (forward pairing). Seethe Materials and methods for further details. (C) Experimental design. Four experiments were performed. In experiment 1, harnessed bees were subjected todifferential olfactory PER conditioning (Train) and then released and tested in the Y-maze (Test). In experiment 2, harnessed bees were subjected to differentialolfactory PER conditioning (Train) and then tested in the same Pavlovian context (Test). In experiment 3, free-walking bees were subjected to differential olfactoryconditioning in the Y-maze (Train) and then harnessed for PER testing (Test). In experiment 4, free-walking bees were subjected to differential olfactoryconditioning in the Y-maze (Train) and then tested in the same context (Test).

950

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

two subgroups, one in which 1-nonanol was the CS+ and 1-octanolthe CS−, and another in which the contingencies were reversed(1-nonanol was the CS− and 1-octanol the CS+). For all threegroups (water, quinine and NaCl) there were no significantdifferences in acquisition according to which odor was the CS+ orthe CS− (two-factor ANOVA for repeated measurements; factorodor; water group: F1,52=0.43, P=0.51; quinine group: F1,47=0.07,P=0.78; NaCl group: F1,53=0.26, P=0.61) so that results werepooled within groups and presented in terms of a CS+ versus CS−discrimination irrespective of odorant identity.Fig. 2A shows the learning performance of three conditioned groups

of bees, which satisfied the criterion established for further testing (seetheMaterials andmethods), namely that they responded correctly to theCS+ in the last conditioning trial.Thesebees also responded correctly tothe CS+ in the previous trial, thus confirming previous findingsshowing that once bees start responding correctly to the CS+ inolfactory PER conditioning, they keep responding to it in further trials(Pamir et al., 2011). All three groups learned to discriminate the CS+from the CS− during conditioning (Fig. 2A; water group: factor CS:F1,53=107.14, P<0.0001; factor interaction: F5,265=23.41, P<0.0001;quinine group: factor CS: F1,48=113.79, P<0.0001; factor interaction:F5,240=25.33, P<0.0001; NaCl group: factor CS: F1,54=429.36,P<0.0001; factor interaction: F5,270=42.21, P<0.0001). However, thediscrimination performance varied depending on the US associatedwith the CS− (factor group: F2,155=4.18, P<0.02). Specifically,although the acquisition curves for the rewarded odor (CS+) wereidentical (F2,155=1.32, P=0.27), the acquisition curves for the non-rewarded odor (CS−) differed significantly (F2,155=5.71, P<0.005). Inother words, the three reinforcers, water, quinine and NaCl, haddifferent capacities to inhibit responses to the CS−. Tukey tests showedthat when taken globally, the CS− curves of the NaCl andwater groupsdiffered significantly (P<0.01) but the CS− curve of the quinine groupdid not differ from that of the water group (P=0.47) or the NaCl group(P=0.94).

Retention in the operant context of the Y-mazeAfter PER conditioning, bees were kept for 2 h in individual glasstubes and then transferred to the Y-maze for two transfer tests in theabsence of reinforcement. In the learning test, the CS+ and the CS−were presented against each other in the new context; in the avoidancetest, the CS− was presented against the novel odor eugenol (Nod) todetermine to what extent the CS− learned in the Pavlovian contextinduced avoidance and thus choice of the novel odor.In the learning test, all three groups showed a tendency to prefer the

previously rewarded CS+ to the CS− because their choice of the CS+

was above 50% (Fig. 2B). A comparison between groups bymeans ofa one-factor ANOVA for independent measurements yielded nodifferences between groups (F2,155=1.87, P=0.16). However,comparing the proportion of choices of the CS+ to a theoreticalvalue of 50% within each group showed that only the NaCl groupexhibited a significant preference with 76% of the bees preferring theCS+ (t54=4.56, P<0.0001), whereas the quinine and the water groupsexhibited a similar non-significant performance with 61% of the beespreferring the CS+ in both cases (water group: t53=1.66, P=0.10;quinine group: t48=1.60, P=0.12). These results confirm the strongerinhibitory effect of 3 mol l−1 NaCl as an aversive reinforcement andthe weaker and comparable inhibitory strength of distilled water and60 mmol l−1 quinine solution in olfactory PER conditioning.

In the avoidance test, all three groups tended to avoid the CS− infavor of the novel odor as their choice of the CS− was below 50%(Fig. 2C). A comparison between groups showed no significantdifferences (F2,155=0.90, P=0.41). However, comparing theproportion of choices of the CS− to a theoretical value of 50%within each group showed that the NaCl group exhibited a higher andsignificant avoidance of the CS− (t54=−3.04, P<0.005) whereas thequinine and water groups showed a non-significant performance(water group: t53=−1.09,P=0.28; quinine group: t48=−1.30,P=0.20).These results thus reaffirm that in olfactory PER conditioning, NaClinduced stronger aversion when paired with a CS− whereas bothwater and quinine had a weaker and similar aversive effect. As aconsequence, both discrimination learning and retention in the mazewere better in the NaCl group owing to the stronger aversive nature ofNaCl as negative reinforcement.

Experiment 2: differential conditioning in the Pavloviancontext of PER conditioning and retention in the samePavlovian contextThis experiment was conceived as a control for experiment 1. Thelevel of correct choices in the Y-maze (CS+ choices in the learningtest) was lower than that observed in PER conditioning. It was thusimportant to determine whether 2 h after conditioning, and afterbeing subjected to the same post-conditioning handling, beesconserved the memories acquired in the olfactory PER conditioningwhen tested in that context.

Acquisition in Pavlovian PER conditioningThe first phase (differential olfactory PER conditioning) wasidentical to that of experiment 1. For all three groups (water,quinine and NaCl) therewere no significant differences in acquisitionaccording to which odor was the CS+ or the CS− (two-factor

CS+ water CS+ quinine CS+ NaClCS– water CS– quinine CS– NaCl

A B Ca

b

b,c

c

Con

ditio

ned

PE

R (%

)

Cho

ice

of C

S+

over

CS

– (%

)

Cho

ice

of C

S–

over

Nod

(%)100

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

01 2 3 4 5 6 Water Quinine NaCl Water Quinine NaCl

Learning trial Learning test Avoidance test

*

*

Fig. 2. Experiment 1: differentialolfactoryPERconditioningandretentionin the operant context of the Y-maze.(A) Acquisition curves during differentialolfactory PER conditioning. NaCl group,N=55; quinine group, N=49; water group,N=54. Different letters indicate significantdifferences between curves. (B) Retentionperformance in the Y-maze; the barsrepresent the percentage of choices of theCS+ in the learning test opposing CS+versus CS−. *P<0.05. (C) Retentionperformance in the Y-maze; the barsrepresent the percentage of choices of theCS− in the avoidance test opposing CS−versus the novel odor (Nod) eugenol.*P<0.05. Dashed horizontal lines indicaterandom choice between test alternatives.

951

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

ANOVA for repeated measurements; factor odor: water group,F1,31=1.56, P=0.22; quinine group, F1,32=0.64, P=0.43; NaCl group,F1,29=0.27, P=0.61) so that results were pooled within groups.As in the previous experiment, all three groups learned to

discriminate the CS+ from the CS− during conditioning (Fig. 3A;water group, factor CS: F1,32=25.06, P<0.0001, factor interaction:F5,160=7.71, P<0.0001; quinine group, factor CS: F1,33=48.67,P<0.0001, factor interaction: F5,165=11.50, P<0.0001; NaCl group,factor CS: F1,30=113.22, P<0.0001, factor interaction: F5,150=24.61,P<0.0001). Performance varied significantly depending on the USassociated with the CS− (F2,95=8.89, P<0.001). Specifically, althoughthe acquisition curves for the rewarded odor (CS+) were identical(F2,95=0.76, P=0.47) between groups, the acquisition curves for thenon-rewarded odor (CS−) differed significantly (F2,95=9.93,P<0.001).When taken globally, the CS− curves of the NaCl and thewater groupsdiffered significantly (Tukey test; P<0.01) whereas the differencesbetween the CS− curve of the quinine group and those of thewater andNaCl groups were marginally non-significant (P=0.05 and P=0.08,respectively). The amount of discrimination reached at the end ofconditioning (CS+ responses minus CS− responses) was the same inexperiments 1 and 2 for the water group (F1,85=1.75, P=0.19), thequinine group (F1,81=0.61, P=0.44) and the NaCl group (F1,84=1.65,P=0.20), thus showing that conditioning yielded the same results inexperiments 1 and 2 for all three groups. In both experiments, NaClwasthe strongest aversive reinforcer, followedbyquinine and thenbywater.

Retention in Pavlovian PER conditioningAfter being placed in individual glass tubes for 2 h as for bees inexperiment 1 and then being transferred to metal tubes for PERtesting, bees were presented with the CS+, the CS− and the novelodor Eugenol (Nod) in a random sequence. Fig. 3B shows that allthree groups responded mainly to the CS+, less to the CS− andpractically not at all to the Nod, thus confirming the lowgeneralization between both CS odors and eugenol. Importantly,the level of responses to the CS+ was similar to that reached at theend of conditioning in all three groups. Thus, the handlingprocedure to which bees were exposed after PER conditioning didnot induce a loss of their memories.No differences were found between groups (two-factor ANOVA

for repeated measures; factor US group: F2,95=0.4, P=0.66), but a

highly significant effect was introduced by the odorant tested (factorodor: F2,190=87.29, P<0.0001). Tukey tests showed that althoughthe NaCl group exhibited significant retention as it discriminated theCS+ from the CS− (Fig. 3B, P<0.0001), the quinine group showedan intermediate significant preference for the CS+ over the CS−(P<0.05) and thewater group a non-significant preference (P=0.74).

In order to refine the analysis of retention performances, wefocused on bees showing a CS-specific memory, i.e. respondingonly to the CS+ and not to the CS− or Nod. Fig. 3C shows that theproportion of bees with CS-specific memory was higher in the NaClgroup, intermediate in the quinine group and lower in the watergroup, consistent with the learning performance. Marascuilo testsfor comparing multiple proportions showed that the CS specificmemory of the NaCl group was significantly higher than that of thewater group (P<0.05); no other comparisons were significant.

Experiment 3: differential conditioning in the operantcontext of the Y-maze and retention in the Pavlovian contextof PER conditioningThis experiment is the reversed version of experiment 1. Bees wereconditioned to discriminate 1-octanol and 1-nonanol in the Y-mazeand then tested under harnessing conditions.

Acquisition in the operant context of the Y-mazeAs in experiments 1 and 2, bees were conditioned over 12 trials todiscriminate theCS+associatedwith sucrose from theCS− associatedeither with water, quinine or NaCl. Within each US group, there wereno significant differences in acquisition according to which odor, 1-octanol or 1-nonanol, was the CS+ or the CS− (water group:F1,9=0.007, P=0.94; quinine group: F1,20=0.009, P=0.93; NaClgroup:F1,19=1.19,P=0.29) so that results were pooled within groups.

Fig. 4 represents the learning performance of the three groups ofbees which efficiently responded to the CS+ in the last visit to themaze. All three groups learned the task as shown by the improvementin the percentage of correct choices at the end of conditioning (watergroup: F11,110=2.57, P<0.01; quinine group: F11,231=2.30, P<0.02;NaCl group: F11,220=2.92, P<0.002). Responses in the maze show atypical variable pattern, which reflects both the information acquiredthrough learning and the expression of exploratory tendencies thatdecrease the level of correct choices.

CS+ water CS+ quinine CS+ NaClCS– water CS– quinine CS– NaCl

A B C

a

a,b

a,b

b

PE

R (%

)

CS

+ sp

ecifi

c m

emor

y

100

80

60

40

20

01 2 3 4 5 6 CS+ CS– Nod Water Quinine NaCl

Learning trial Test US group

WaterQuinineNaCl

aa

a

a

b

bb

b

a,b

a

b

0.6

0.5

0.4

0.3

0.2

0.1

0

Fig. 3. Experiment 2: differential olfactory PER conditioning and retention in the same context. (A) Acquisition curves during differential olfactory PERconditioning; white symbols: NaCl group, N=31; quinine group, N=34; water group, N=33. Different letters indicate significant differences between curves.(B) PER responses upon stimulation with the CS+, the CS− and the novel odor (Nod) eugenol. The bars represent the percentage of PER responses.(C) Retention performance expressed as CS-specific memory levels, i.e. the proportion of bees responding only to the CS+ and neither to the CS− nor to the Nod.Different letters indicate significant differences between curves.

952

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

A comparison between groups revealed significant differences(factor group: F2,51=3.75, P<0.05) and a global significant increaseof correct responses (factor trial: F11,561=5.83, P<0.0001) but nosignificant interaction (F22,561=1.00, P=0.46). When taken globally,the curves of the NaCl and the water groups differed significantly(Tukey test; P<0.05) while the differences between the curves of thequinine group and those of the water and the NaCl groups were notsignificant (P=0.07 and P=0.89, respectively). Thus, in the operantcontext of the Y-maze, the effect of quinine is more similar to that ofNaCl which remains the strongest aversive reinforcement.

Retention in Pavlovian PER conditioningAfter the last visit to the maze, bees were placed in individual glasstubes and then fixed in metal tubes for PER testing upon odorstimulation. Beeswere presentedwith theCS+, theCS− and the novel

odorant Eugenol (Nod). Surprisingly, no bee responded to eitherodorant in these transfer tests. The absence of response transfer mayreflect amemory loss 2 h after conditioning or a refusal (or incapacity)to express the existingmemories under the new restrictive, harnessingconditions. The latter option would be consistent with an increase ofstress levels following immobilization, which would preclude beesresponding appropriately to the learned odors. To test this hypothesis,beeswere trained in theY-maze and tested 2 h after conditioning in thesame maze after similar handling and timing as in experiment 3.

Experiment 4: differential conditioning in the operantcontext of the Y-maze and retention in the same context ofthe Y-mazeAcquisition in the operant context of the Y-mazeBees were conditioned in the Y-maze as in experiment 3. Withineach US group, there were no significant differences in acquisitionaccording towhich odor, 1-octanol or 1-nonanol, was the CS+ or theCS− (water group: F1,6=0.21, P=0.66; quinine group: F1,11=3.62,P=0.08; NaCl group: F1,12=2.00, P=0.18) so that results werepooled within groups. All three groups of bees, which efficientlyresponded to the CS+ in the last visit to the maze, learned the task(Fig. 5A; water group: F11,77=2.25, P<0.02; quinine group:F11,132=1.87, P<0.05; NaCl group: F11,143=2.13, P<0.03) butdiffered significantly in their performance (factor group:F2,32=14.77, P<0.001). Interestingly, the curves of the NaCl andquinine groups did not differ significantly (P=0.99), thus showingthat both reinforcements induced similar learning performanceswhen paired with the CS−. However, both groups differedsignificantly from the water group (quinine versus water:P<0.001; NaCl versus water: P<0.001), which induced less-efficient learning when paired with the CS−. These results thusconfirm the tendencies observed in experiment 3 and show that inthe Y-maze, the effect of quinine is similar to that of NaCl anddistinct from that induced by water.

Retention in the operant context of the Y-mazeAfter conditioning, bees were placed in individual glass tubes wherethey were kept for 2 h. Thereafter, they were replaced individually inthe Y-maze where they were subjected to a learning test opposingthe CS+ and the CS− previously learned and to an avoidance testopposing the CS− and the Nod.

CS+ sucrose CS– water aa,bb

Cor

rect

cho

ices

(%)

100

80

60

40

20

01 2 3 4 5 6

Trial7 8 9 10 11 12

CS+ sucrose CS– quinineCS+ sucrose CS– NaCl

Fig. 4. Experiment 3: differential conditioning in the operant context ofthe Y-maze and retention in the context of PER conditioning. Acquisitioncurves (% correct choices) during differential olfactory conditioning in theY-maze; NaCl group, N=21; quinine group, N=22; water group, N=11. Thecurves of the NaCl and the water groups differed significantly (P<0.05) whilethe differences between the curves of the quinine group and those of the waterand the NaCl groups were marginally non-significant (P=0.07) and non-significant (P=0.89), respectively. Once harnessed in the contention tubes,no bees responded to either odorant. The dashed horizontal line indicatesrandom choice between maze alternatives.

CS+ sucrose CS– water

Cor

rect

cho

ices

(%)

100

80

60

40

20

01 2 3 4 5 6

Trial7 8 9 10 11 12

CS+ sucrose CS– quinineCS+ sucrose CS– NaCl

A B C

Cho

ices

of C

S+

over

CS

– (%

)

100

80

60

40

20

0

Cho

ices

of C

S–

over

Nod

(%)

100

80

60

40

20

0Water Quinine NaCl Water Quinine NaCl

**

**

abb

Learning test Avoidance test

Fig. 5. Experiment 4: differential conditioning in the operant context of the Y-maze and retention in samemaze context. (A) Acquisition curves (% correctchoices) duringdifferential olfactoryconditioning in theY-maze;NaCl group,N=14; quininegroup,N=13;watergroup,N=8.Thecurvesof theNaClandquininegroupsdidnot differ significantly (P=0.99). Yet, both groups differedsignificantly from thewater group (P<0.001). (B)Retention performance in theY-maze; the bars representthe percentage of choices of the CS+ in the learning test opposing CS+ versus CS−. *P<0.05. (C) Retention performance in the Y-maze; the bars represent thepercentage of choices of the CS− in the avoidance test opposing CS− versus the novel odor eugenol. *P<0.05. The dashed horizontal lines indicate random choicebetween maze alternatives.

953

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

In the learning test, all three groups showed a tendency to preferthe previously rewarded CS+ to the CS− (Fig. 5B) as shown by thefact that their CS+-choice level was above of 50% in all cases. Nodifferences between groups were detected (F2,32=0.23, P=0.80).However, comparing within each group the proportion of CS+choices to 50% showed that both the quinine and the NaCl groupsexhibited a significant preference with 85% and 86%, respectively,of the bees preferring the CS+ (quinine group: t12=3.32, P<0.01;NaCl group: t13=3.68, P<0.005). On the contrary, the bias towardsthe CS+ observed in the water group was not significant (t6=1.16,P=0.29). These results show that, consistently with their learningperformances, the quinine and the NaCl groups had better retentionthan the water group. Such an effect confirms the stronger inhibitoryeffect of both quinine and NaCl 3 mol l−1 as aversivereinforcements and the weaker strength of distilled water in thecontext of the Y-maze.In the avoidance test, only the quinine and the NaCl groups

tended to avoid the CS− in favor of the novel odor while the watergroup chose randomly between both odors (Fig. 5C). A comparisonbetween groups was close to significance (F2,32=2.70, P=0.08).Comparing within each group the proportion of choices of the CS−to a theoretical value of 50% showed that, while the water group hada non-significant performance (t7=0,P=1), the quinine and the NaClgroups avoided significantly the CS− (quinine group: t12=5.50,P<0.001; NaCl group: t13=2.51, P<0.05). These results thusreaffirm that in the operant context of the Y-maze, quinine andNaCl had a strong and similar inhibitory effect.

DISCUSSIONOur results show that the impact of some compounds used as negativereinforcements in learning experiments with honey bee varies with theexperimental context chosen to train the animals. In both the Pavloviancontext and the operant context of the Y-maze, the solution of3 mol l−1 NaCl induced better olfactory discrimination, robust mid-term memories and efficient transfer of these memories from thePavlovian to the operant context. Also, in both the Pavlovian and theoperant contexts, distilled water induced less olfactory discrimination,less retention and no or lessmemory transfer from the Pavlovian to theoperant context. However, a clear variation was found for the solutionof 60 mmol l−1 quinine. In olfactory PER conditioning, quinineinduced intermediate acquisition and discrimination between theCS+ and the CS− (Figs 2A and 3A), as well as intermediate memoryretention (Fig. 3B,C). After PERconditioning, it induced only a partialand non-significant memory transfer to the operant context ofthe Y-maze (Fig. 2B,C). Conversely, in the operant context of theY-maze, quinine had a more powerful effect, similar to that of NaCland definitely stronger than that of distilled water (Fig. 5). In theY-maze, it supported efficient discrimination learning (Figs 4 and 5A)and retention (Fig. 5B,C). These results thus show that quinine doesnot have the same impact act as a negative reinforcement in twodifferent experimental contexts in which the common taskwas to learnto discriminate between the same two odorants.

Negative reinforcements in olfactory PER conditioningExperiments 1 and 2 allowed us to compare the impact of differentsubstances as negative reinforcements in Pavlovian olfactory PERconditioning. In this protocol, pairing distilled water, quinine or NaClwith a CS− induced different discrimination learning from a CS+paired with sucrose solution (Figs 2A and 3A). Whereas 3 mol l−1

NaCl clearly improved odor discrimination thus revealing a strongeraversive effect, distilled water induced a lower level of discriminationconsistent with its less-aversive nature; 60 mmol l−1 quinine induced

an intermediate discrimination level. The results of the transfer tests inthe Y-maze (experiment 1; see Fig. 2B,C) reflected those obtained inolfactory PER conditioning because only NaCl induced significantCS+ preference and CS− avoidance, thus confirming its strongeraversive nature. Water and quinine induced non-significant biases,thus revealing that the bitter substance had a lower impact because itspairing with a CS− did not enhance olfactory discrimination from aCS+ associated with sucrose solution. These results were confirmedby experiment 2 where the pattern of acquisition was the same as inexperiment 1 and where retention tests showed again a higher levelof specific memory in the NaCl group, a significantly lower level ofspecific memory in the water group and an intermediate levelof specificmemory in the quinine group (Fig. 3C), consistent with theresults obtained in the acquisition phase of the experiment. Note thatin experiments 1 and 2, the stimulation time with the negativereinforcements was the same for all bees (6 s; see the Materials andmethods) so that differences in performance cannot be attributed todifferences in US exposure.

The fact that quinine was not experienced as an intense aversivereinforcement accounts for previous results showing that bees in thesame conditions as those in experiments 1 and 2, i.e. harnessed inindividual metal tubes, ingest without reluctance different kinds ofnoxious substances, including bitter ones, even if they die as aconsequence of the malaise induced by this ingestion (Ayestaránet al., 2010). Clearly, for these bees, quinine solution (10 and100 mmol l−1) and other pure bitter substances (e.g. salicin,amygdaline, L-canavanine) are not penalizing enough to inducerejection when delivered to the mouth parts even if 2 h afteringestion, some of these substances (e.g. quinine) induce high levelsof mortality (50% and 40% for quinine solutions of 10 and100 mmol l−1, respectively).

Taken together, experiments 1 and 2 show that different USinduce different memories because of their different aversiveeffects. For harnessed bees subjected to olfactory PER conditioning,quinine solution was not the strongest aversive reinforcementbecause it yielded only intermediate levels of acquisition, retentionand specific memory. NaCl, by contrast, was the most aversive USreinforcement as it induced better discrimination, retention andspecific memory levels.

Negative reinforcements in olfactory learning in the Y-mazeIn the context of the Y-maze, the quinine solution used as a negativereinforcement was definitely a more-efficient aversive US thandistilled water (Fig. 5A) and induced acquisition and retention levelscomparable to those induced by the NaCl solution (Fig. 5B,C). Inthis case, the hypothesis that a bitter compound acts as a powerfulaversive stimulus was proved, in contrast to results observed inolfactory PER conditioning.

Contrary to PER conditioning, and as a result of the operantframework of experiments 3 and 4, it was not possible to ensure thatthe time of exposure to the negative reinforcement was the same inall three groups. In the maze, the bees control their US exposure.Yet, the learning curve of the water group was always lower thanthose of the NaCl and the quinine groups which ran parallel to eachother; this means that water bees made more errors that broughtthem more frequently to water than the quinine and the NaCl beesdid with their respective negative reinforcements. This, however,did not make water a more powerful negative US. The fact that inthis protocol quinine was as strong as NaCl cannot be explained by ahigher number of errors in the quinine group (i.e. by more contactswith the quinine solution) because the NaCl and the quinine groupsexhibited a similar number of errors throughout acquisition.

954

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

Our results thus confirm findings from experiments in which free-flying bees were trained to discriminate a visual target paired withsucrose solution from one or various visual distracters paired withquinine solution, under the assumption that quinine would act as astrong penalty favoring learning and discrimination (Chittka et al.,2003; Dyer and Chittka, 2004; Avargues-Weber et al., 2010; Reseret al., 2012; Avargues-Weber and Giurfa, 2014). In our experiments,bees did not fly within the small Y-maze, but as in the works justmentioned, they could move, explore the maze and compare theodor stimuli and their corresponding outcomes through sampling.This capacity to move and actively express choice behavior, notgranted by olfactory PER conditioning, seems crucial to favordiscrimination. However, it may also make the quantification of thisdiscrimination difficult because bees could show a decrease of CS+choices, not only because of a learning or memory deficit, butpotentially because of exploratory behavior in the maze. This effectmay be particularly strong in a transfer test from PER conditioningto the maze: if, after harnessing, the animal recovers the capacity tomove and explore the novel environment of the maze, it may retrieveits olfactory memories, leading them to the CS+ but, at the sametime, feel inclined to visit both arms of the maze. These opposingtendencies might explain why the level of correct choices decreasesfrom PER conditioning to the test in theY-maze, and affects both theappetitive and the aversive memories.

Information transfer between different experimentalcontextsOur results show that memory transfer was possible from PERconditioning to the Y-maze, but not in the opposite direction. Themain difference between both experimental contexts resides in thefreedom to move granted by the Y-maze but not by the contentiontubes. Olfactory PER conditioning is a Pavlovian protocol supportingCS−US (odor-gustatory reinforcement) associations, over which thebehavior of the conditioned bee has no control (Bitterman et al.,1983). Olfactory discrimination learning in the Y-maze also supportsPavlovian associations as the trained bee learns associations betweenodors and gustatory reinforcements; yet, in this case, the choicebehavior of the insect is crucial and determines the reinforcementoutcome experienced effectively. Thus, while both Pavlovian andoperant associations are formed in the Y-maze, the latter aredeterminant for the animal to perform correctly in this context.In our experiments, only bees that succeeded in learning the CS+

effectively in their respective training contexts were subjected totransfer tests. Differences in transfer performance were not due,therefore, to differences in CS+ learning, but only to the way thebees evaluated the CS− penalty. Differences in the aversive strengthof CS− reinforcers, together with exploratory behaviors (see above),account for the transfer performances observable in theY-maze afterPER conditioning. In this case, the bee that recovers the freedom toexpress the contents of its memory behaviorally shows odordiscrimination whenever the CS− reinforcement is strong enough(as in the case of NaCl) to overcome the potential obscuring effectof exploratory behavior.The situation is, however, different when the animal loses

behavioral freedom by being transferred from the Y-maze to theindividual harnesses for testing PER. In this case, no PER responsecould be recorded, irrespective of the CS− reinforcers used forconditioning. Olfactory memories were not lost during the 2 hperiod elapsed since the end of training in the maze and the transferto the tubes for PER testing, as shown by experiment 4. However,they were not expressed in contention, in the form of PER. The factthat the same result was found for all three CS− reinforcers shows

that the absence of responses was a general effect, due to the novelsituation imposed on the bees. Harnessing bees in tubes implies aloss of motor freedom and requires the expression of a behavior,PER, that was not necessarily contingent with sucrose in the maze.Thus, while recovering freedom by being transferred from PERconditioning to the maze may be experienced as a positive situation,the reverse situation, from maze conditioning to immobilization inthe tubes, may be experienced as a negative, stressful situationinhibiting memory expression.

Various studies have shown that the olfactory experience acquiredby bees in a context can affect their subsequent behavior in a differentcontext (e.g. Gerber et al., 1996; Sandoz et al., 2000; Chaffiol et al.,2005; Mc Cabe and Farina, 2009). However, none of these previousstudies provided the precise control of experience and/or thebi-directional transfer in both contexts achieved in our work. Bothfeatures suggest that the constraints imposed by the Y-maze and bythe contention tubes in our work constitute different challenges for abee transferred from one context to the other. While a transfer fromthe contention tubes to the maze may be experienced as a positivesituation as a result of the recovery of freedom of movement, the lossof this freedom in the reverse transfer from the maze to the contentiontubes may be accompanied by high stress levels that need to bereduced before observing the expression of memories acquired in theoperant context. These higher levels of stress may trigger a series ofphysiological processes underlying a change in perception of quininestrength. For instance, stress activates an opioid-like system in the bee(Núñez et al., 1997; Balderrama et al., 2002), which leads to apotential increase of tolerance to harmful stimuli, without affectingthe response to appetitive stimuli (Balderrama et al., 2002). Underthese conditions, tolerance to the effects of quinine, either pre- (taste)or post-ingestive (malaise) might be increased.

Mechanisms underlying changes in reinforcement strengthbetween contextsThe concentrated NaCl solution was the strongest aversivereinforcement in both contexts. This was probably due to the factthat concentrated saline solutions disrupt osmotic equilibrium andlead rapidly to death. In the case of the NaCl solution used in ourexperiments (3 mol l−1), ingestion of 20 µl induces 80%mortality 2 hafter ingestion (Ayestarán et al., 2010), thus showing its lethal effects.Distilled water, by contrast, was theweakest aversive reinforcement inboth contexts. This is not surprising because water is a neutralsubstance with no dramatic impact on the physiology of the bees.Quinine solution was the reinforcement whose strength variedbetween the Pavlovian and the operant contexts. What, besidesstress-associated mechanisms, could underlie such variation?

Another possible mechanism of aversive-strength regulation reliesonNPY-type peptides, which play a prominent role inmanagement ofstress responses and emotion in mammals (Bannon et al., 2000;Thorsell et al., 2000; Thorsell and Heilig, 2002). In Drosophila,neuropeptide F (NPF), the counterpart of the mammalianneuropeptide Y (NPY), also promotes resilience to diverse stressorsand prevents aggressive behavior (Wu et al., 2003,Wu et al., 2005a,b).In particular, NPFR1, a G-protein-coupled NPF receptor, exerts aninhibitory effect on larval aversion to diverse stressful stimuli. Thissystem also regulates the propensity to feed on potentially toxic foodin Drosophila larvae (Wu et al., 2005a); larvae are more prone tofeed on foods adulterated with 0.5% quininewith longer deprivationand stress periods (Wu et al., 2005a) and this risk-prone feeding isassociated with higher levels of expression of NPFR1, whose up-regulation is sufficient to trigger intake of noxious food in non-deprived larvae. Conversely, disruption of neural NPFR1 signaling

955

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

in deprived animals leads to a decrease in noxious-food feeding (Wuet al., 2005a). In the honey bee, two NPY-related genes wereidentified, npf and snpf (Hummon et al., 2006), but a receptor foronly the short (s) peptide (sNPFR) was found (Hauser et al., 2006);sNPFR is up-regulated by food deprivation, which may constitute aform of stress (Ament et al., 2011). We thus suggest that theexperimental constraints imposed by the contention tubes of thePavlovian preparation may result in a stress-induced up-regulationof sNPFR, which in turn will promote intake of quinine solution.

PerspectivesSo far, we have no proof concerning the existence of specificreceptors for bitter taste at the peripheral level in the gustativesystem of the bee (de Brito Sanchez et al., 2005, 2014; de BritoSanchez, 2011). The present results suggest that a peripheraldetection of quinine during learning in the Y-maze is possible, andwould be a fruitful line of investigation for further research.Indeed, odor discrimination is fast in this context so that post-ingestive malaise would not have time to develop. Low mortalitywas observed in our experiments (<5%), thus suggesting that thebees of the quinine group did not ingest the quinine but based theiravoidance on brief taste contacts, as has previously been observedin experiments with free-flying honeybees (Avargues-Weber et al.,2010). These contacts could have occurred at the beginning of thetraining phase and may be sufficient to prevent the bees fromchoosing the CS− odorant. The honey bee genome (HoneybeeGenome Sequencing Consortium, 2006) showed that in thereduced set of gustatory receptor genes of this insect, nohomolog of the bitter-taste receptor genes of Drosophila couldbe found (Robertson and Wanner, 2006). Thus, the mechanismsunderlying peripheral detection of pure quinine solution remain tobe determined.Our study shows that some gustatory processes can be modulated

by experience. This is an important conclusion that not only haspractical consequences for the design of experiments involving tasteinput, but also has more general consequences, as it underlines thatperception is not absolute but may also be subjected to top-downmodifications depending upon experience and context. Furtherstudies should focus on the mechanisms underlying suchmodulation by exploring, among others, the hypotheses raised inour work.

MATERIALS AND METHODSHoney bee (Apis mellifera Linnaeus 1758) foragers, from a hive located50 m from the laboratory, were caught at the hive in the morning. They wereplaced in glass vials, and cooled on ice until they stopped moving.

StimuliIn both the olfactory conditioning of PER and in the Y-maze, bees weretrained using a differential conditioning procedure to discriminate twoodorants, one of which (CS+) was always paired with 1 mol l−1sucrosesolution, while the other (CS−) was paired either with distilled water(control group), 60 mmol l−1 quinine solution (bitter group) or 3 mol l−1

NaCl (NaCl group). We thus aimed at determining whether the presence of aputative bitter reinforcement (quinine) on the CS− enhanced olfactorydiscrimination. Quinine (60 mmol l−1) was chosen as it is highlyconcentrated and improves visual discrimination of small colordifferences in free-flying bees (Avargues-Weber et al., 2010). Quininesolutions of lower (10 mmol l−1) and higher concentrations (100 mmol l−1)induce high mortality when ingested by harnessed bees (Ayestarán et al.,2010) so the solution used in our experiments was potentially noxious to thebees if imbibed. NaCl (3 mol l−1) was chosen because its association with aCS− increases discrimination performances in olfactory PER conditioning(Getz et al., 1986).

The odorants used for conditioning were 1-nonanol and 1-octanol, whichare hardly discriminable from each other in olfactory PER conditioning,when one of them is paired with sucrose and the other with absence of it(Guerrieri et al., 2005). Eugenol, was used as a novel odorant in the retentiontests. It was chosen because it is easily discriminable from 1-octanol and1-nonanol and induces low generalization from the conditioned odors(Guerrieri et al., 2005). All chemicals were obtained from Sigma-Aldrich(Saint Quentin Fallavier, France).

In both the olfactory conditioning of PER and in the Y-maze, the CS+wasalways paired with sucrose solution 1 M. The CS− was paired either withdistilled water (control group), 60 mmol l−1 quinine solution (bitter group)or 3 mol l−1NaCl (NaCl group). We thus aimed to determine whether thepresence of a putative bitter reinforcement (quinine) on the CS– enhancedolfactory discrimination. Experiments were fully balanced, i.e. independentgroups were trained either with 1-nonanol as CS+ and 1-octanol as CS− orwith the reverse contingency. An independent group was used for eachnegative reinforcement (water, quinine, NaCl) so that six groups wereconditioned in total.

Pavlovian olfactory conditioning of PER (experiments 1 and 2)Bees were individually harnessed in small tubes so that they could only movetheir antennae and mouthparts, including the proboscis (Fig. 1A). They werethen fed with 2 µl of 1 mol l−1 sucrose solution and kept in the dark and inhigh humidity for approximately 2 h (Matsumoto et al., 2012). Fifteenminutes before conditioning, each subject was checked for intact PER bylightly touching the antennae with a toothpick imbibed with 1 mol l−1

sucrose solution without subsequent feeding. Hungry, motivated beesrespond to this stimulation by extending reflexively the proboscis to lick thesucrose. Extension of the proboscis beyond the virtual line between the openmandibles was counted as a response (Fig. 1A). Animals that did not exhibita PER at this stage were not used in the experiments (<5%).

Odors were delivered by means of an olfactometer (Fig. 1A), which sent aconstant clean-air stream in which odor pulses of known duration could beintroduced. Four µl of each odor were applied onto a filter paper placedwithin a syringe connected to the odor-delivery setup. The air stream wasproduced by an air pump (Rena Air 400, Annecy, France) and directed to therelevant syringes by means of electronic valves (Lee Company S.A.,Voisins-le-Bretonneux, France) controlled by a computer. In the absence ofolfactory stimulation, the air stream passed through a syringe containing aclean piece of filter paper (clean air stream). During olfactory stimulation, theair streamwas directed to a syringe containing a filter paper loadedwith odor.After 6 s stimulation, the air stream was again redirected to the odorlesssyringe until the next olfactory stimulation. The whole setup was placed infront of an air extractor, which impeded the accumulation of residual odorsafter delivery of an olfactory stimulus.

Conditioning (experiments 1 and 2) consisted of 6 CS+ and 6 CS− trials,presented in a pseudo-random sequence starting with the CS+ or the CS−odor in a balanced way. The intertrial interval was 10 min. Eachconditioning trial lasted 30 s. A trial started when a harnessed bee wasplaced between the olfactometer and the air extractor for 10 s to allowfamiliarization with the training situation. Thereafter, the CS was deliveredfor 6 s. Three seconds after CS onset, the antennae and the proboscis werestimulated for 6 s with a toothpick soaked in the US solution (sucrose in CS+trials, and water, quinine or NaCl in CS− trials). If PER occurred, the beewas allowed to feed during this stimulation period. Otherwise, the probosciswas extended by means of the toothpick, and the solution was brought intodirect contact with the proboscis for the same period. In all cases weobserved that the drop at the tip of the toothpick disappeared, suggesting thatthe amount imbibed was similar for all solutions. The interstimulus intervalwas 3 s. The beewas left in the conditioning place for 11 s and then removed.

During each trial, we recorded whether the bee extended its proboscisafter CS onset and before US onset (conditioned response). We thenestablished and represented the percentage of bees exhibiting a conditionedresponse to the CS+ and the CS− during conditioning trials.

For retention tests under harnessing conditions (experiments 2 and 3),bees were presented with the CS+ and the CS− odorants used duringolfactory PER conditioning as well as with the novel odor eugenol in orderto assess the specificity of the retrieved olfactory memories (Matsumoto

956

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

et al., 2012). Eugenol is easily discriminable from 1-octanol and 1-nonanolused for conditioning so that generalization between conditioned odors andthe novel odor was negligible (Guerrieri et al., 2005). Odorant stimulationwas identical to that of conditioning trials (6 s) but no US was given. Odorpresentations were spaced by 10 min. The sequence of odors variedrandomly between bees. No changes were detected depending on odorsequence. After the tests, PER to the sucrose US was checked once again.Animals unable to show PER at this point (<1%) were not considered for theanalyses.

Olfactory conditioning in the Y-maze (experiments 3 and 4)Each bee was confined in an individual glass tube where it could move andplaced afterwards twice in the Y-maze deprived of odors (Fig. 1B) tofamiliarize it with the set-up (pre-training) (Carcaud et al., 2009). A smalldrop of sucrose solution was placed in the decision zone of the maze, at theintersection of both arms, to incite the bee to further explore the set-up. Aftergetting the sucrose reward, the bee was returned to its tube, the filter papercovering the floor of the maze replaced with a clean one.

The odorants and reinforcements used to condition the bees were thesame as for olfactory PER conditioning (see above). Originally, the acrylicY-mazewas positioned under homogeneous red light (Carcaud et al., 2009),provided by a cold light source in a dark room, which prevented the beesfrom using visual cues for orientation and from trying to fly. However,preliminary experiments (not shown) demonstrated that identicalperformances were obtained in the laboratory under normal room light.Thus, to avoid a change from dark to light between the Y-maze and the PERconditioning, experiments were performed under room light.

The entrance channel and the arms of the maze were 1.9 cm high, and8 cm and 6 cm long, respectively. The arms were at a 90 deg angle, each at135 deg from the entrance channel. The maze was placed on a rectangularbase (13.5×14.5 cm) from which it could be removed to be cleaned withethanol. The maze was covered by a glass plate (10×15 cm). The floor of themaze was covered by a piece of filter paper, which was replaced by a cleanone after each visit of a honeybee to the maze.

The entrance to each arm was defined at its narrowest point, connectingthe arms to the entrance channel (see dotted lines, Fig. 1B). In each arm, amicropipette tip containing a piece of filter paper (1×20 mm) loaded with4 μl odor substance was inserted into a hole in the floor. The tips weresealed at their bottom and covered with a plastic net hood at their top toavoid direct contact with the chemicals. Each tip was placed 1.5 cm fromthe arm entrance, so that honeybees entering an arm experienced the odoremanating from it. An air stream (15 ml min−1) filtered by active charcoalwas humidified and driven from the back of each arm by means of plastictubes. This allowed the odors to be driven towards the decision area of themaze. The tips were renewed during the experiment approximately everyhour, or if the bee walked on one of them. The glass plate covering themaze impeded bees from escaping and allowed better concentration ofodors.

One and a half µl of each US solution was placed on a plastic discpositioned close to the back wall of each arm and at 3 cm of the odor tips. Inthis way, a bee entering an arm of the maze perceived first the odor and thenthe reinforcing solution, thus creating the conditions for a forward pairingbetween odor and its associated reinforcement.

Conditioning consisted of 12 training visits (12 trials) to the Y-mazepresenting the conditioned odors. The bee selected for training could freelywalk and choose between the odorants presented in the arms of the maze.Between visits, the bee was returned to its glass tube, the Y-shaped filterpaper covering the floor of the maze was changed and the glass plate and themaze cleaned with alcohol. Special care was taken to always eliminate allpossible traces of alcohol that could affect the bee’s choice.

As in the olfactory PER conditioning, the CS+ was always paired with1 mol l−1 sucrose solution whilst the CS− was paired either with distilledwater (control group), 60 mmol l−1 quinine solution (bitter group) or3 mol l−1 NaCl (NaCl group). The position of the CS+ and the CS− and oftheir associated reinforcers was interchanged between arms following arandom sequence in order to avoid the development of side biases.

During each trial, we noted the first arm chosen by the bee. The firstchoice could be correct, i.e. choice of the arm with the CS+ leading to

sucrose solution or incorrect, i.e. choice of the arm with the CS− leading tothe negative reinforcement. If the choice was correct, the entrance of thenegative arm was immediately blocked by means of a plastic wall to impedethe bee experiencing the negative reinforcement. If the choice was incorrect,the bee was free to move to the arm with the CS+, and the negative arm wasthen blocked as explained above. Once the bee drank the sucrose solutionand went back to the decision chamber, it was captured and brought backagain to its glass tube until the next conditioning trial.

Two retention tests separated by 10 min were performed: in one test(learning test), the CS+ and the CS− odors previously conditioned werepresented against each other; in the other test (avoidance test), the CS− odorwas presented against a novel odor which was eugenol as in retention testsfollowing PER conditioning (see above). The sequence of tests variedrandomly between bees and no changes were detected depending on testorder. No reinforcement was provided during the tests. The learning testallowed measuring retention and transfer of the information learned in thePavlovian context to the operant context; the avoidance test allowed todetermine to what extent the CS− learned in the Pavlovian context inducedavoidance and thus choice of the novel odor eugenol.

From one bee to the next, the placement of the test odors was swappedbetween arms, so that no effect of the sides could influence the results.Between tests, the maze was cleaned with ethanol and a new filter paper wasplaced to cover the floor. For every bee and test, we recorded the first choicewithin the Y-maze. Afterwards, the bee was removed from the maze.

Transfer between experimental contextsAt the end of the acquisition phase of olfactory PER conditioning, bees wereleft in their contention tubes for 23 h in an oven at 30°C and then releasedinto individual glass tubes where they could move. From there, they weretransferred, one by one, into the Y-maze for olfactory testing (experiment 1).In the case of experiment 2, which implied testing in harnessing conditions,bees were handled in the same way except that they were maintained in thecontention tubes all the time.

At the end of the acquisition phase in the Y-maze, bees were individuallyharnessed in contention tubes and placed in a warmed oven at 30°C for2 h. The bees were subsequently brought individually to the olfactometer(Fig. 1A) for PER testing (experiment 3). In the case of experiment 4, whichimplied retesting in the Y-maze, bees were handled in the same mannerexcept that at the end of the 2 h rest period, they were released into individualglass tubes from which they could be re-transferred to the Y-maze. The 2 hperiod between the two phases of each experiment was chosen as itcorresponds to the minimal resting period that is commonly used in PERconditioning to habituate bees to the harnessing conditions before the start ofconditioning (Matsumoto et al., 2012).

StatisticsPER to odors (conditioned response) was quantified as a dichotomousvariable (1: yes; 0: no) and expressed as the percentage of animals exhibitingPER to an olfactory stimulation. For the retention/transfer tests afterPavlovian PER conditioning, we used only bees that responded correctly totheCS+ in the last conditioning trial, irrespective of their response to theCS−.In a differential conditioning, learning results from the excitatory responseinduced by the CS+ (here the odor paired with sucrose solution) and theinhibitory response induced by the CS− (here the odor paired with eitherwater, quinine or NaCl), both leading to discrimination of the CS+ from theCS−. As our question was how the US associated with the CS− (NaCl, wateror quinine) modulates learning (i.e. if a certain US was more effective toinduce inhibitory learning of the CS−), it was necessary to keep the CS+response level fixed to focus exclusively on the impact of the US paired withthe CS−. This is why only efficient CS+ learners with potentially differentCS− experiences were brought to the retention/transfer tests.

Acquisition performances were analyzed by means of analyses ofvariance (ANOVAs) for repeated measurements both for within andbetween-group comparisons. Monte Carlo studies have shown that it ispermissible to use ANOVA on dichotomous data under controlledconditions, which are met by our experiments (equal cell frequencies andat least 20 degrees of freedom of the error) (Lunney, 1970). Post hoccomparisons were performed using Tukey tests.

957

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

For retention tests, the proportions of responses to the three test odorspresented (CS+, CS− and novel odor eugenol) were analyzed by means ofANOVA of repeated measures. In these retention tests, besides quantifyingthe proportion of responses/choices, we established the proportion of beesexhibiting a CS-specific memory, which was defined as the proportion ofbees responding only to the CS+ and neither to the CS− nor the Nod.Multiple comparisons between CS specific memory levels were performedby means of Marascuilo tests, which allow comparisons of multipleproportions. All statistical analyses were carried out using Statistica 5.5(StatSoft, France), and differences were considered significant if the P valuewas smaller than 0.05.

Odor choice in the Y-maze during training was quantified as adichotomous variable (1: correct; 0: incorrect) and represented as thepercentage of animals exhibiting correct responses along the 12conditioning trials. For retention/transfer tests, we used only bees thatchose the CS+ correctly and not the CS− in the last visit to the maze. In thisway, only efficient CS+ learners with potentially different CS− experienceswere brought to the retention/transfer tests.

Acquisition performances in the Y-maze were analyzed by means ofanalyses of variance (ANOVAs) for repeated measurements both for withinand between-group comparisons. For retention tests in the Y-maze(experiments 1 and 4), we quantified the proportion of choices in eachdual choice test (CS+ vs CS− and CS− vs Nod). Thus, a single value per beeper test was obtained. To avoid pseudo-replication, a one-sample t-test wasused to test the null hypothesis of this proportion not being different from thetheoretical value of 50% (random choice). Between-group comparisonswere performed by means of a one-factor ANOVA for independentmeasures.

AcknowledgementsWe thank two anonymous reviewers for useful comments and corrections toprevious versions of the manuscript. We also thank Lucie Hotier for valuablebeekeeping support during our experiments.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsM.G.d.B.S. conceived and designed the experiments; M.S. performed theexperiments; M.G.d.B.S., M.S. and M.G. analyzed the data; M.G.d.B.S., M.S.,A.A.-W., A.G.D. and M.G. discussed the data. M.G.d.B.S. andM.G. wrote the paper.

FundingWe thank the French Research Council (CNRS) and the University Paul Sabatier forsupport. M.G. acknowledges the generous support of the Institut Universitaire deFrance (IUF) and of the French National Research Agency (ANR; projectPHEROMOD). A.G.D. is funded by the Australian Research Council [fellowshipDP0878968].

ReferencesAment, S. A., Velarde, R. A., Kolodkin, M. H., Moyse, D. and Robinson, G. E.(2011). Neuropeptide Y-like signalling and nutritionally mediated gene expressionand behaviour in the honey bee. Insect Mol. Biol. 20, 335-345.

Avargues-Weber, A. and Giurfa, M. (2014). Cognitive components of color visionin honey bees: how conditioning variables modulate color learning anddiscrimination. J. Comp. Physiol. A 200, 449-461.

Avargues-Weber, A., de Brito Sanchez, M. G., Giurfa, M. and Dyer, A. G. (2010).Aversive reinforcement improves visual discrimination learning in free-flyinghoneybees. PLoS ONE 5, e15370.

Avargues-Weber, A., Deisig, N. and Giurfa, M. (2011). Visual cognition in socialinsects. Annu. Rev. Entomol. 56, 423-443.

Ayestaran, A., Giurfa, M. and de Brito Sanchez, M. G. (2010). Toxic but drank:gustatory aversive compounds induce post-ingestional malaise in harnessedhoneybees. PLoS ONE 5, e15000.

Balderrama, N., Nun ez, J. A., Guerrieri, F. and Giurfa, M. (2002). Differentfunctions of two alarm substances in the honeybee. J. Comp. Physiol. A 188,485-491.

Bannon, A.W., Seda, J., Carmouche, M., Francis, J. M., Norman, M. H., Karbon,B. and McCaleb, M. L. (2000). Behavioral characterization of neuropeptide Yknockout mice. Brain Res. 868, 79-87.

Bitterman, M. E., Menzel, R., Fietz, A. and Schafer, S. (1983). Classicalconditioning of proboscis extension in honeybees (Apis mellifera). J. Comp.Psychol. 97, 107-119.

Carcaud, J., Roussel, E., Giurfa, M. and Sandoz, J.-C. (2009). Odour aversionafter olfactory conditioning of the sting extension reflex in honeybees. J. Exp. Biol.212, 620-626.

Chaffiol, A., Laloi, D. and Pham-Delegue, M. H. (2005). Prior classical olfactoryconditioning improves odour-cued flight orientation of honey bees in awind tunnel.J. Exp. Biol. 208, 3731-3737.

Chittka, L., Dyer, A. G., Bock, F. and Dornhaus, A. (2003). Psychophysics: beestrade off foraging speed for accuracy. Nature 424, 388.

Darling, F. M. C. and Slotnick, B. M. (1994). Odor-cued taste avoidance: a simpleand efficient method for assessing olfactory detection, discrimination andmemoryin the rat. Physiol. Behav. 55, 817-822.

deBrito Sanchez, M. G. (2011). Taste perception in honey bees.Chem. Senses 36,675-692.

deBrito Sanchez, M. G., Giurfa, M., de PaulaMota, T. R. andGauthier, M. (2005).Electrophysiological and behavioural characterization of gustatory responses toantennal ‘bitter’ taste in honeybees. Eur. J. Neurosci. 22, 3161-3170.

deBrito Sanchez, M. G., Ortigao-Farias, J. R., Gauthier, M., Liu, F. L. andGiurfa,M. (2007). Taste perception in honeybees: just a taste of honey? Arthropod PlantInteract. 1, 69-76.

de Brito Sanchez, M. G., Lorenzo, E., Su, S., Liu, F., Zhan, Y. and Giurfa, M.(2014). The tarsal taste of honey bees: behavioral and electrophysiologicalanalyses. Front. Behav. Neurosci. 8, 1-16.

Dyer, A. G. and Chittka, L. (2004). Bumblebees (Bombus terrestris) sacrificeforaging speed to solve difficult colour discrimination tasks. J. Comp. Physiol. A190, 759-763.

Erber, J., Masuhr, T. H. and Menzel, R. (1980). Localization of short-term memoryin the brain of the bee, Apis mellifera. Physiol. Entomol. 5, 343-358.

Gerber, B. andHendel, T. (2006). Outcome expectations drive learned behaviour inlarval Drosophila. Proc. R. Soc. B Biol. Sci. 273, 2965-2968.

Gerber, B., Geberzahn, N., Hellstern, F., Klein, J., Kowalksy, O., Wuestenberg,D. and Menzel, R. (1996). Honey bees transfer olfactory memories establishedduring flower visits to a proboscis extension paradigm in the laboratory. Anim.Behav. 52, 1079-1085.

Getz, W. M., Bruckner, D. and Smith, K. B. (1986). Conditioning honeybees todiscriminate between heritable odors from full and half sisters. J. Comp. Physiol. A159, 251-256.

Giurfa, M. (2007). Behavioral and neural analysis of associative learning in thehoneybee: a taste from the magic well. J. Comp. Physiol. A 193, 801-824.

Giurfa, M. and Sandoz, J.-C. (2012). Invertebrate learning and memory: fifty yearsof olfactory conditioning of the proboscis extension response in honeybees. LearnMem. 19, 54-66.

Glendinning, J. I., Brow, H., Capoor, M., Davis, A., Gbedemah, A. and Long, E.(2001). A peripheral mechanism for behavioral adaptation to specific “bitter” tastestimuli in an insect. J. Neurosci. 21, 3688-3696.

Glendinning, J. I., Davis, A. and Ramaswamy, S. (2002). Contribution of differenttaste cells and signaling pathways to the discrimination of “bitter” taste stimuli byan insect. J. Neurosci. 22, 7281-7287.

Guerrieri, F., Lachnit, H., Gerber, B. and Giurfa, M. (2005). Olfactory blocking andodorant similarity in the honeybee. Learn. Mem. 12, 86-95.

Hauser, F., Cazzamali, G., Williamson, M., Blenau, W. and Grimmelikhuijzen,C. J. P. (2006). A review of neurohormone GPCRs present in the fruitflyDrosophila melanogaster and the honey bee Apis mellifera. Progr. Neurobiol. 80,1-19.

Honeybee Genome Sequencing Consortium, Weinstock, G. M., Robinson,G. E., Gibbs, R. A., Weinstock, G. M., Weinstock, G. M., Robinson, G. E.,Worley, K. C., Evans, J. D., Maleszka, R. et al. (2006). Insights into social insectsfrom the genome of the honeybee Apis mellifera. Nature 443, 931-949.

Hummon, A. B., Richmond, T. A., Verleyen, P., Baggerman, G., Huybrechts, J.,Ewing, M. A., Vierstraete, E., Rodriguez-Zas, S. L., Schoofs, L., Robinson,G. E. et al. (2006). From the genome to the proteome: uncovering peptides in theApis brain. Science 314, 647-649.

Ito, E., Kobayashi, S., Kojima, S., Sadamoto, H. and Hatakeyama, D. (1999).Associative learning in the pond snail, Lymnaea stagnalis. Zool. Sci. 16, 711-723.

Kemenes, I., O’Shea, M. and Benjamin, P. R. (2011). Different circuit andmonoamine mechanisms consolidate long-term memory in aversive and rewardclassical conditioning. Eur. J. Neurosci. 33, 143-152.

Laska, M. and Metzker, K. (1998). Food avoidance learning in squirrel monkeysand common marmosets. Learn Mem. 5, 193-203

Lunney, G. H. (1970). Using analysis of variance with a dichotomous dependentvariable: an empirical study. J. Educ. Meas. 7, 263-269.

Matsumoto, Y., Menzel, R., Sandoz, J.-C. and Giurfa, M. (2012). Revisitingolfactory classical conditioning of the proboscis extension response in honeybees: a step toward standardized procedures. J. Neurosci. Meths. 211, 159-167.

Mc Cabe, S. I. and Farina, W. M. (2009). Odor information transfer in the stinglessbee Melipona quadrifasciata: effect of in-hive experiences on classicalconditioning of proboscis extension. J. Comp. Physiol. A 195, 113-122.

Menzel, R. (1999). Memory dynamics in the honeybee. J. Comp. Physiol. A 185,323-340.

Menzel, R. and Giurfa, M. (2001). Cognitive architecture of a mini-brain: thehoneybee. Trends Cognit. Sci. 5, 62-71.

958

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology

Menzel, R., Erber, J. and Masuhr, T. (1974). Learning and memory in thehoneybee. In Experimental Analysis of Insect Behaviour (ed. L. Barton-Browne),pp. 195-217. Berlin: Springer.

Nun ez, J., Almeida, L., Balderrama, N. and Giurfa, M. (1997). Alarm pheromoneinduces stress analgesia via an opioid system in the honeybee. Physiol. Behav.63, 75-80.

Pamir, E., Chakroborty, N. K., Stollhoff, N., Gehring, K. B., Antemann, V.,Morgenstern, L., Felsenberg, J., Eisenhardt, D., Menzel, R. and Nawrot, M. P.(2011). Average group behavior does not represent individual behavior in classicalconditioning of the honeybee. Learn Mem. 18, 733-741.

Reilly, S. and Schachtman, T. R. (2008). Conditioned Taste Aversion: Neural andBehavioral Processes. New York: Oxford University Press.

Reser, D. H., Wijesekara Witharanage, R., Rosa, M. G. P. and Dyer, A. G. (2012).Honeybees (Apis mellifera) learn color discriminations via differential conditioningindependent of long wavelength (green) photoreceptor modulation. PLoS ONE 7,e48577.

Robertson, H. M. andWanner, K.W. (2006). The chemoreceptor superfamily in thehoney bee, Apis mellifera: expansion of the odorant, but not gustatory, receptorfamily. Genome Res. 16, 1395-1403.

Rodrıguez-Girones, M. A., Trillo, A. and Corcobado, G. (2013). Long term effectsof aversive reinforcement on colour discrimination learning in free-flyingbumblebees. PLoS ONE 8, e71551.

Salloum, A., Colson, V. and Marion-Poll, F. (2011). Appetitive and aversivelearning in Spodoptera littoralis larvae. Chem. Senses 36, 725-731.

Sandoz, J. C., Laloi, D., Odoux, J. F. and Pham-Delegue, M. H. (2000). Olfactoryinformation transfer in the honeybee: compared efficiency of classicalconditioning and early exposure. Anim. Behav. 59, 1025-1034.

Scott, K. (2004). The sweet and the bitter of mammalian taste. Curr. Opin.Neurobiol. 14, 423-427.

Takeda, K. (1961). Classical conditioned response in the honey bee. J. Insect.Physiol. 6, 168-179.

Tanimura, T. and Kikuchi, T. (1972). Response to bitter substances in Drosophilamelanogaster. Jpn. J. Gen. 47, 376-377.

Thorsell, A. and Heilig, M. (2002). Diverse functions of neuropeptide Y revealedusing genetically modified animals. Neuropeptides 36, 182-193.

Thorsell, A., Michalkiewicz, M., Dumont, Y., Quirion, R., Caberlotto, L.,Rimondini, R., Mathe, A. A. and Heilig, M. (2000). Behavioral insensitivity torestraint stress, absent fear suppression of behavior and impaired spatial learningin transgenic rats with hippocampal neuropeptide Y overexpression. Proc. Natl.Acad. Sci. USA 97, 12852-12857.

Weiss, L. A., Dahanukar, A., Kwon, J. Y., Banerjee, D. and Carlson, J. R. (2011).The molecular and cellular basis of bitter taste inDrosophila.Neuron 69, 258-272.

Wu, Q., Wen, T., Lee, G., Park, J. H., Cai, H. N. and Shen, P. (2003).Developmental control of foraging and social behavior by the Drosophilaneuropeptide Y-like system. Neuron 39, 147-161.

Wu, Q., Zhao, Z. and Shen, P. (2005a). Regulation of aversion to noxious food byDrosophila neuropeptide Y- and insulin-like systems. Nat. Neurosci. 8,1350-1355.

Wu, Q., Zhang, Y., Xu, J. and Shen, P. (2005b). Regulation of hunger-drivenbehaviors by neural ribosomal S6 kinase inDrosophila.Proc. Natl. Acad. Sci. USA102, 13289-13294.

Yamamoto, T., Shimura, T., Sako, N., Yasoshima, Y. and Sakai, N. (1994). Neuralsubstrates for conditioned taste aversion in the rat.Behav. Brain Res. 65, 123-137.

Yarmolinsky, D. A., Zuker, C. S. and Ryba, N. J. P. (2009). Common sense abouttaste: from mammals to insects. Cell 139, 234-244.

959

RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, 949-959 doi:10.1242/jeb.117333

TheJournal

ofEx

perim

entalB

iology